Archive via cron take a lot of time


Our instance of matomo is taking a lot of time to process archive via cron. The last one took 34105s (~9.5h) plus 850s of scheduled task
The issue seems coming from the processing of segments.

We’re are running on matomo 3.13.5 with a mysql db and a dedicated server to do the archive.

For the year 2020 so far the whole matomo instance is processing 519 website for a total of 33 306 930 visits / 358 388 368 pageviews.

Here the command line used to process:
console core:archive - - url=https://our_dedicated_server - - concurrent-requests-per-website=9

Is there a way to improve the duration of the processing?


Any hint on my issue?